EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

ISFJ-RAG: Interventional Suppression of Hallucinations via Counter-Factual Joint Decoding Retrieval-Augment Generation

Yuezhao Liu, Wei Li, Yijie Wang, Ningtong Chen, Min Chen

Big Data and Cognitive Computing · 09.02.2026 · 10 (2) · 56

DOI: 10.3390/bdcc10020056 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Big Data and Cognitive Computing, online erschienen am 09.02.2026. Verfasst von Yuezhao Liu und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 41 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Jiangxi Provincial Natural Science Foundation, Nanchang University Innovation Fund Project.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Although retrieval-augmented generation (RAG) technology mitigates the hallucination issue in large language models (LLMs) by incorporating external knowledge, and combining reasoning models can further enhance RAG system performance, retrieval noise and attention bias still lead to the diffusion of factual errors in problems such as factual queries, multi-hop questions, and unanswerable questions. Existing methods struggle to effectively suppress “high-confidence hallucinations” in long-chain reasoning due to their failure to decouple knowledge bias effects from causal reasoning paths. To address this, this paper proposes the ISFJ-RAG framework, which dynamically intervenes in hallucinations through counterfactual joint decoding. First, a structural causal model (SCM) reveals three root causes of hallucinations in RAG systems: irrelevant knowledge interference, reasoning path bias, and spurious correlations in self-attention mechanisms. A dual-decoder architecture is further designed: the total causal effect decoder models the global relationship between user queries and knowledge, while the knowledge bias effect decoder captures potential biases induced by external knowledge. A dynamic modulation module converts the latter’s output into a proxy measure of hallucination bias. By computing individual treatment effects (ITEs), the bias component is removed from the full generation distribution, achieving simultaneous suppression of knowledge-irrelevant and reasoning-irrelevant hallucinations. Ablation experiments validate the robustness of average token log-probability as a confidence metric. Experiments demonstrate that on the RAGEval benchmark, ISFJ-RAG improves generation completeness to 86.89% (+5.49%) while reducing hallucination rates to 10.39% (−2.5%) and irrelevance rates to 4.44% (−2.99%).

Förderung

  • National Natural Science Foundation of China
  • Jiangxi Provincial Natural Science Foundation
  • Nanchang University Innovation Fund Project

Zitieren

APA

Liu, Y., Li, W., Wang, Y., Chen, N., & Chen, M. (2026). ISFJ-RAG: Interventional Suppression of Hallucinations via Counter-Factual Joint Decoding Retrieval-Augment Generation. Big Data and Cognitive Computing, 10(2), 56. https://doi.org/10.3390/bdcc10020056

Vancouver

Liu Y, Li W, Wang Y, Chen N, Chen M. ISFJ-RAG: Interventional Suppression of Hallucinations via Counter-Factual Joint Decoding Retrieval-Augment Generation. Big Data and Cognitive Computing. 2026;10(2):56. doi:10.3390/bdcc10020056

BibTeX
@article{Liu2026,
  author = {Liu, Yuezhao and Li, Wei and Wang, Yijie and Chen, Ningtong and Chen, Min},
  title = {ISFJ-RAG: Interventional Suppression of Hallucinations via Counter-Factual Joint Decoding Retrieval-Augment Generation},
  journal = {Big Data and Cognitive Computing},
  year = {2026},
  volume = {10},
  number = {2},
  pages = {56},
  doi = {10.3390/bdcc10020056}
}
Kostenlos starten →